Automated blood pressure recorders are used with increasing frequency by pregnant women, mostly without proper evaluation of their accuracy. We compared blood pressures (BP) recorded by 2 automated noninvasive devices, the Spacelabs 90207 ambulatory blood pressure monitor and the OMRON HEM 705 CP portable self-initiated device, with blood pressures recorded by routine sphygmomanometry in 79 pregnant women either considered 'at risk' for preeclampsia or with mild hypertension in pregnancy. The Spacelabs device tended to overestimate systolic BP by a mean 11 (SD=8) mmHg and diastolic BP by 5 (SD=7) mmHg for phase 5 pressure (p<0.001) but was similar to routine BPs for diastolic phase 4 pressures. The OMRON device tended to underestimate diastolic (phase 4) pressure by 4 (SD=6) mmHg (p<0.001) but gave similar systolic and diastolic (phase 5) pressures to routine sphygmomanometry. However, for both devices there was considerable individual patient variability in accuracy. When using these devices to record a limited number of blood pressure recordings, as in this study, we suggest that individual comparison with mercury sphygmomanometry be made in each pregnant woman before accepting the validity of these recordings.